As retina specialists, our primary objective in managing Retinopathy of Prematurity (ROP) has historically been the immediate preservation of vision—halting VEGF-driven neovascularization and preventing retinal detachment. To this end, treatments such as pan-retinal photocoagulation (laser) and intravitreal anti-VEGF therapies (like bevacizumab) have been revolutionary. However, the successful clinical regression of acute ROP does not equate to the complete normalization of retinal architecture.
Physiological foveal maturation is an exquisitely precise process requiring the centrifugal displacement of the inner retinal layers and the concurrent regression of the central retinal vasculature during late gestation and early postnatal life, reaching adult configurations around 45 months. Disruption of this process by premature birth and ROP results in subtle, persistent deviations in macular microvascular development that can last into adulthood.
A highly comprehensive 2026 systematic review and meta-analysis by Mollapour ballujeh et al., published in the Survey of Ophthalmology, offers an unprecedented quantitative synthesis of optical coherence tomography (OCT) and OCT angiography (OCTA) parameters across 31 studies (encompassing 2,786 eyes). By comparing term-born infants, preterm infants without ROP, and various ROP subgroups (laser-treated, anti-VEGF-treated, spontaneously regressed, and untreated), this data challenges our understanding of long-term ROP sequelae. Here is a deep dive into the structural and microvascular legacies of ROP and what they mean for your long-term clinical management of these patients.
Arrested Foveal Pruning: The Superficial Capillary Plexus (SCP)
One of the most profound takeaways from the meta-analysis is the stark difference in the development of the foveal superficial capillary plexus (SCP) between full-term infants and those with a history of ROP.
Normal foveal maturation requires central capillaries to regress. In eyes with a history of ROP, this vascular pruning is arrested. The meta-analysis revealed that term infants exhibit significantly lower foveal SCP vessel density (VD) compared to all ROP subgroups. Specifically, the mean difference (MD) in foveal SCP VD for term infants was -9.29% (P < 0.001) when compared to laser-treated eyes, -9.09% against bevacizumab-treated eyes, and -7.78% when compared to eyes with spontaneously regressed ROP.
Importantly, the difference between term-born infants and preterm infants without ROP was much smaller (MD: -1.85%) and not initially statistically significant, meaning that prematurity alone accounts for less than one-quarter of the elevated foveal VD seen in ROP. The persistence of the SCP within the foveal center is primarily driven by the disease’s pathogenic angiogenic activity rather than a simple interruption of gestation.
Interestingly, the choice of therapy leaves a distinct quantitative signature. When comparing anti-VEGF therapy to spontaneous regression, bevacizumab-treated eyes exhibited a significantly higher foveal SCP VD (MD: 4.62%, P = 0.008). This intermediate quantitative phenotype suggests that while anti-VEGF therapy spares the peripheral retina and is theoretically less destructive, it results in only partial preservation of physiological central remodeling, not a return to true baseline.
Peripheral Plasticity: The Parafoveal SCP
While the foveal center demonstrates arrested development, the superficial parafoveal region tells a story of remarkable postnatal plasticity.
In the parafovea, the density relationship flips: term infants demonstrated a higher parafoveal SCP VD compared to laser-treated (MD: 3.77%, P < 0.001) and bevacizumab-treated (MD: 4.26%, P < 0.001) groups. However, there was no significant difference between term infants and those with spontaneously regressed ROP (MD: 0.94%, P = 0.186).
This suggests that without the destructive intervention of laser or the potent cytokine alteration of anti-VEGF, the superficial parafoveal vascular network retains a robust capacity for adaptive remodeling and can approach near-normal physiological states. The superficial parafoveal network appears highly flexible and forgiving over time, setting up a sharp contrast with the deeper retinal layers.
The Deep Capillary Plexus (DCP): A Biomarker of Long-Term Vulnerability
For the retina specialist, the deep capillary plexus (DCP) findings are perhaps the most clinically alarming. The DCP is tasked with supplying the highly metabolically demanding outer retina, and disruptions here signify a lasting limitation in the patient’s long-term metabolic reserve.
Much like the superficial layer, the foveal DCP VD was significantly lower in term infants compared to all other groups, including laser (MD: -8.09%), bevacizumab (MD: -7.48%), and regressed ROP (MD: -7.30%). However, the critical vulnerability is found in the parafoveal DCP.
The meta-analysis revealed that spontaneously regressed ROP eyes exhibited a significantly lower parafoveal DCP VD compared to preterm infants without ROP (MD: -3.03%, P < 0.001). While a 3% reduction might sound numerically minor compared to the massive shifts seen in the fovea, this effect was incredibly consistent across the literature. It indicates that developmental angiogenesis in the deep plexus has a severely limited capacity for postnatal recovery once it has been disrupted by ROP.
This dissociation between the behavior of the superficial plexus (which shows plasticity) and the deep plexus (which shows permanent deficits) provides a mechanistic framework for understanding why these children often exhibit subtle, persistent visual and contrast sensitivity deficits later in life despite absolute clinical quiescence of the disease. Children with reduced parafoveal DCP VD harbor a reduced microvascular reserve that is completely invisible on standard fundus examination.
Foveal Avascular Zone (FAZ) Topography
The Foveal Avascular Zone (FAZ) metrics perfectly corroborate the vessel density findings. Enlargement of the FAZ during late gestation normally accompanies the near-complete loss of central capillaries. Because this process is halted in ROP, FAZ dimensions are universally constricted in this population.
Term infants consistently exhibited a larger FAZ compared to active ROP (MD: 0.157 mm²), laser-treated (MD: 0.175 mm²), regressed (MD: 0.092 mm²), and untreated ROP (MD: 0.164 mm²) groups. Because a reduction of 0.1 to 0.2 mm² implies substantial retention of intrafoveal vasculature, it is clear that the presence of the disease itself—not just the therapeutic modality chosen—drives the primary deviation from physiological development. The failure of the FAZ to expand is a permanent structural footprint of ROP.
Structural Remodeling: Foveal Thickness and Inner Retinal Retention
Optical coherence tomography (OCT) structural data provides the anatomical correlate to the OCTA vascular findings. Centrifugal displacement of the inner retinal layers is necessary to form the normal foveal depression. In ROP, these layers are frequently retained.
The quantitative meta-analysis showed that laser-treated eyes had a remarkably greater foveal thickness compared to spontaneously regressed eyes (MD: 35.5 µm, P < 0.001) and untreated ROP eyes (MD: 26.6 µm, P = 0.004). An increase of over 25 to 35 µm vastly exceeds normal inter-individual variability and strongly suggests the aggressive retention of inner retinal layers, potentially exacerbated by an inflammatory response, post-laser tissue remodeling, or subclinical edema.
This structural anomaly is not benign. Prior quantitative analyses in the literature demonstrate that every 10–15 µm increase in foveal thickness is associated with a measurable, progressive decline in visual acuity in children with a history of ROP. Even in cases of spontaneous regression, foveal development proceeds at a slower pace and carries long-term functional consequences.
Methodological Considerations and Caveats
While this meta-analysis presents a robust and highly powered overview of ROP microvasculature, retina specialists must contextualize these findings within the limitations of current imaging technology. The included studies were predominantly retrospective and utilized a variety of commercial spectral-domain and swept-source OCTA devices (e.g., Optovue RTVue XR Avanti, Zeiss Cirrus AngioPlex, Topcon DRI Triton). Differences in automated segmentation boundaries, varying image quality thresholds, and susceptibility to motion artifacts in pediatric populations can introduce between-study variability.
Furthermore, sensitivity analyses within the meta-analysis revealed that extreme prematurity (birth weights < 1500g) itself confers a distinct microvascular signature independent of ROP. Therefore, evaluating a patient’s microvascular reserve requires a heavily weight-stratified interpretation.
Re-Conceptualizing ROP Management
For decades, ROP has been viewed primarily as an acute neonatal emergency. Once the ridge has flattened and plus disease has resolved, the condition is often considered “cured.” However, the data synthesized by Mollapour ballujeh et al. mandates a paradigm shift.
ROP must be conceptualized as a chronic developmental disorder with durable microvascular signatures that persist well beyond clinical regression. The substantial retention of foveal superficial capillaries, the failure of FAZ expansion, and the permanent deficit in the metabolically critical deep capillary plexus collectively prove that these retinas are not physiologically normal.
What does this mean for the clinical retina specialist?
- Look Beyond the Macula’s Surface: A visually “flat” and clinically quiet macula on standard biomicroscopy may harbor significant inner retinal retention and a compromised deep vascular bed.
- Embrace OCTA as a Biomarker: Region- and layer-specific OCTA metrics (particularly foveal SCP VD and parafoveal DCP VD) should be integrated into long-term surveillance strategies. These parameters have the potential to serve as powerful biomarkers to stratify patients at highest risk for subtle, late-onset visual and contrast sensitivity deficits.
- Refine Treatment Expectations: While anti-VEGF therapies preserve peripheral retina better than laser ablation, we must counsel parents (and temper our own expectations) that intravitreal injections do not permit the fovea to achieve a fully normal physiological state. The fovea remains a permanently altered structure.
As our premature patients grow into adulthood, they carry the microvascular legacy of their neonatal disease. By integrating advanced OCT and OCTA imaging into our standard longitudinal follow-up, we can transition from simply preventing childhood blindness to proactively managing the lifelong functional visual potential of ROP survivors.
Ballujeh, Masoumeh Mollapour, et al. “Optical coherence tomography and angiography for the assessment of retinal microvasculature in retinopathy of prematurity patients: A systematic review and meta-analysis.” Surv. Ophthalmol., vol. S0039-6257, no. 26, 9 July 2026, pp. 00095-00090, doi:10.1016/j.survophthal.2026.07.005.
